Why wind needs feed-in tariffs (and why it is not the enemy of nuclear)
Posted by Jerome a Paris on March 4, 2008 - 12:30am in The Oil Drum: Europe
Topic: Economics/Finance
Tags: nuclear, power prices, wind [list all tags]
| An argument often heard against wind is that it costs a lot in public subsidies for a solution that will always have a limited
impact (because it still produces only a small fraction of overall needs, and because of its unreliability linked to its
intermitten nature). This is an argument worth addressing in detail, especially when it is pointed out, as the graph shows, that
wind is already almost competitive with the other main sources of electricity, which suggests that it might not even need
the subsidies then (and the increase in commodity prices since that graph was prepared using 2004 data, only reinforces that
argument).
|  |
We are on the brink of a new energy order. Over the next few decades, our reserves of oil will start to run out and it is imperative that governments in both producing and consuming nations prepare now for that time. We should not cling to crude down to the last drop - we should leave oil before it leaves us. That means new approaches must be found soon.
The above, from an article by Fatih Birol, the increasingly strident chief economist of the International Energy Agency, suggests that we need to develop all non-carbon based energy sources as quickly as we can to avoid the coming energy crunch from oil depletion. He suggests to push nuclear energy, but that may not be enough - and, as I will show below, the best way to push nuclear is also the best way to promote wind power...
:: ::
Now, if you look at the graph above, it is very easy to see that the long term cost components of wind power and gas power are very different. Nuclear is quite similar to wind in that respect (more, in fact than the above graph suggests), and coal is quite similar to gas.
Wind turbines, once built, generate almost free electricity - they require only some basic maintenance and servicing. That means that they have a marginal cost of production close to zero (ie each additional kWh of production only requires more wind, but no actual spending); that also means that their main long term cost is the repayment of the initial construction cost, in the form of debt repayment and return on capital for the investors.
This has two simple consequences:
- the cost of wind power is essentially set at the time of construction, when the parameters of the financing of the initial investment are agreed, in the form of debt service plus a set return, over an agreed period of time, typically 15-20 years. That cost is fixed and will not vary in accordance with the price at which electricity is actually sold.
- once installed, wind power will always be dispatched - with its negligible marginal cost of production, it will always be cheaper than alternatives, and the only reason not to take such free power will be technical constraints from the network (which I'll discuss later). When dispatched, wind power will move the dispatch curve, and ensure that the marginal cost of production required at that point ot satisfy demand will be lower than if wind power were not available - ie wind power displaces the most expensive power source that would have been needed otherwise, typically a gas-fired plant.
The first consequence noted above is a bit more subtle and needs to be discussed in more detail.
As noted, wind power has high fixed costs, while gas power has low fixed costs but higher variable costs - the cost of procuring fuel. At a time of steadily increasing gas prices, that might seem like an advantage for wind, but, in fact, it is not. The reason for that is that, in today's liberalised markets whereby electricity prices are driven by the marginal cost of production, power prices tend to follow that of gas, since the marginal producer is usually a gas-fired plant. Thus, the variability of gas prices is mirrored in electricity prices, and a gas-fired plant does not really see its competitive position in the market change.
On the other hand, a wind farm, with its fixed costs, makes a lot of money when gas prices (and thus electricity prices) are high, but stands to lose money should at any point electricity prices come down again. The short term profitability of wind farms is driven by factors totally outside of their control (gas prices, which are themselves driven, in the medium term, by oil prices). Should that short term profitability be negative for too long, that can spell trouble for the investment (ie bank loans might be in default - even if temporarily - and the investors then stand to lose the project to the banks. And if that's too likely to happen, banks simply won't lend, because any default (even a temporary one) causes losses and headaches. Essentially, investors and banks must bet that gas prices will stay high enough every single one of the next 15 years for the project to avoid trouble.
To express things differently, the competitiveness of a wind farm - decided at the time of investment - depends on how low the gas prices might go over the next 15 years, whereas the competitiveness of a gas-fired plant depends mostly on the existing power plants - to know the plant's position on the dispatch curve, and thus its likely use. To a much lesser extent, the relative variations of gas and coal prices will also play a role, but this has a second-order impact on revenues.
In short, a gas-fired plant presents a much lower risk profile at the time of the investment, in the sense that the risk of catastrophic loss (from long term price movements) is much less, and that the somewhat higher short term price risk is easier to manage (and financial markets are happy to provide their services there).
That different risk profile is, of course, the reason why wind power needs to be supported in some way by public authorities: markets, left to themselves, will invest in the very technologies (gas and coal) that are the source of all our worries, founded or nor, on the energy front: climate change (coming from carbon emissions), and security of supply (coming from the likely depletion of resources in the long term and the perceived unreliability of suppliers like Russia in the short term).
And the public authority has an actual incentive to encourage wind farms: the long term fixed nature of its price structure presents an unsurmountable risk for the private sector, but it does embed very real value for any entity able to bet on the very long term: a guarantee that prices will be no higher than that fixed cost, whatever the price of oil, in 20 years' time. The markets, except for very specific cases (energy intensive industrialists that know their energy needs in the long term, are not necessarily concerned about temporary interruptions and value long term average prices rather than short term ones), are currently unable to give a value to what is effectively a very long term option on electricity prices - but that value is there.
We know we'll still need electricity in the next 20 years; public policy that works to provide a cap to how high the price of that electricity can go sounds like smart policy - and smart politics.
In fact, on the basis of the value of that option, it can be argued that feed-in tariffs, which provide a stable, guaranteed price to wind power and thus allow the relevant investment to be made with the high-probability perspective of a decent return , are not a subsidy, but a fair transaction, whereby the public authority purchases the guarantee of capped prices in the future in exchange for somewhat higher prices today. The exemple of Denmark quoted above, and the current trends for oil prices, suggest that this is a transaction likely to be highly profitable in the long run, in fact, and thus not at all a subsidy.
Tax credits, as provided in the USA, are a similarly effective mechanism, as they provide a guaranteed minimum income to wind farms and thus ensure that the minimum long term power price threshhold required to make the investment in a wind farm a sensible one is much lower than it would otherwise be, and thus that such investments can be made today - and indeed they are, as the current boom in windpower in the US shows. And the cost-benefit analysis is likely to be similar once wind reaches a sufficient penetration in the market.
A third mechanism that would work as well is NOT the green certificate market regulations used in a few countries (the UK, Australia, Italy), but would rather consist in authorising public authorities to provide financing to the power sector. Given that the main cost of a wind turbine is the fixed financing cost, if you loser the aplicable interest rate and/or required return on capital, you also lower the long term cost of production. Public authorities can borrow money a lot more cheaply, and over much longer periods, than private sector entities, so the cost difference can be quite significant - it can halve the cost for nuclear plant, for instance. And they would not even need to actually provide funds, as this could take the form of payment guarantees. Thus, the public entity would bear that risk of periods of low power prices in exchange, once again, for having a growing portion of power generation coming from carbon-free, capped-cost sources. and the beauty of such guarantees is that they can be provided to all power sources (ie including gas and coal fired plants) in order to avoid the accusation of distorting competition: the cost impact is a lot bigger on wind or nuclear than on gas or coal, and thus the investment decisions will be correspondingly influenced. Charging a flat fee for such a guarantee would make the mechanism transparent and "fair."
The lesson from all this is that wind power does not need subsidies if you make it possible to take into account long term perspectives rather than short term risks. And the same argument applies to nuclear power, so the two technologies are perfectly aligned in that respect - one could even argue that mechanisms that allow to take into account the long term cost/benefit analysis would boost nuclear even more, given that nuclear power plants present the additional risk, from a private investor's perspective, that it is a huge discrete investment, ie it is hard ot invest a small amount in a nuclear plant, you need to sink at least a couple billion euros. Wind farms can at least come in chunks of a few million a piece but, with nuclear, you need to bet big each time, and very few private sector players can afford to concentrate their risks like this.
Of course, this discussion has not even discussed the fact that most existing technologies other than wind are heavily subsidized, either directly, or because they do not have to pay for the externalities they cause. The most obvious example being the lack of price paid, until emissions trading actually comes into force, for the carbon dioxide emissions from gas- or coal-fired plants, or the direct subsidies paid to coal mining in many countries.
At this stage, nuclear advocates might agree with my points and conclude that we need to focus on building nuclear plants, given that wind, being unreliable and small-scale, can never "do the job."
I'd argue that, while personally favorable to nuclear, it's not the easiest solution to deploy in many countries. Given that the State will always bear the ultimate risk for very long term waste management, for catastrophic accident insurance (both impossible to price by the private sector) and for overall safety and security regulation, and that price "support" as proposed above further implicates public authorities, my position is that nuclear power should be run by publicly-owned entities - the EDF model. Under such a model, nuclear can indeed provide a large chunk of our electricity needs.
But even in countries where this model can be applied, there should be no limitation to the development of wind power, and no need for nuclear advocates to demean or mock wind power. Given that it is essentially the same regulatory framework that favors both technologies (with specific regulatory requirements for waste on the one side, and for network reinforcement on the other), they are objective allies in the public debate on energy.
For this *kind* of article I found it well balanced. It was not "anti" but "positive" on both wind and nuclear. I'm curious. Did you take into account actual cost for KWs based on Winds actual output or did you use faceplate capacity instead?
David Walters
left-atomics.blogspot.com
at this level of discussion I have to assume anyone writing is using actual production instead of nameplate capacity :) The chart at the top measures in MWhe.
Here is the actual production from 3 wind farms in Ontario for July 2006 (2007 numbers not out yet):
http://www.ieso.ca/imoweb/marketdata/genDisclosure.asp
GENERATOR DISCLOSURE REPORT - July 2006
Station Name
Station MCR
Capability Factor %
Capability Factor %
Energy Production (MWh)
Production Factor %
Type
Farm Limited Partnership
Development Corporation
For the total output wind is a mere 0.19%:
15%
More Turbines Needed
For Ontario to get to the target of 15% of total output would mean 78 TIMES more turbines than it has now. Two things become important then. The cost to get the number of turbines needed to get a viable output. And the time it would take to build them. At the current rate of construction it would take more than 100 years to build the turbines needed to get to 15%, to build in a more reasonable time would mean ramping up 10 to 20 TIMES the current rate of construction.
The costs and benifits in this article does not include the problems faced with wind spikes, and any efforts needed to smooth out the output.
I've been following the daily output from Wind, here is what I have so far:
We had a low pressure system move through on the higher outputs. Once a high pressure system settles in the output drops by 80-90%.
Additional info on Ontario's wind:
http://www.ieso.ca/imoweb/pubs/marketReports/ORO_Report-2007-2-2.pdf
Adding in the total Name Plate Capacity of 475 Mw into my chart:
The complete daily output from wind can be found here
http://www.ieso.ca/imoweb/pubs/marketReports/download/HourlyWindGen_2008...
I compiled the data from
http://www.ieso.ca/imoweb/pubs/marketReports/download/HourlyWindGen_2008...
To get the number of hits per hour for each percent of name plate for all the wind output from 1-Mar-06 to 26-feb-08, 17,473 measurements.
I get this graph
This shows clearly where the ouput from windturbines spends its time as a percent of the Name Plate. 30% of the hourly hits is above the 27% capacity factor. 70% of the time the hourly output is below 27%. 50% of the time the output is at 13% of Name Plate or less. 5% of the time there is no output at all. 90% of the hits are at or below 50% Name Plate.
David
I used a graph from the Economist, which is itself based on IEA work (dating back from 2004), simply to have a basis for discussion, and to show that current prices per kWh are roughly in the same range.
I've seen numerous studies with various estimates of prices for kWh from various sources, inclusing some that try to include carbon costs (increasing the cost of coal and gas) and balancing costs (increasing that of wind).
This one (from Emerging Energy Research) is probably pretty close to my own experience and estimates for the US (with Europe seeing wind more expensive and nuclear less):
My experience of actual projects I've financed is that wind power will typically cost 3-6c/kWh to produce before taking into account the return on capital (ie profit) - ie long term O&M and debt service costs, depending on wind conditions and the date of construction (recent projects tend to have seen prices increase, like all other technologies, due to commodity prices and supply chain bottlenecks).
EDF's costs at in the 2-3c/kWh for nuclear, but costs for new investment, based on private sector financing, are likely to be significantly higher.
Bonjour!
Une question, Jerome...
Do you know why solar thermal (CSP large scale) is almost never included in those studies ?
CSP seems to be very attractive in independent studies, and a worthwhile (if not totally necessary) addition to any future energy mix ?
Apologies, Jerome, in the course of a long thread I overlooked this reply to my question.
I do feel though that the use of 2004 data considerably understates windpower costs, and perhaps more importantly can falsely give the impression of a more-or-less smooth transition to lower costs.
My real objection though is to off-shore wind power, where costs seem ludicrous - £66bn for 33GW nameplate, around 10-11GW actual average hourly, and that does not include transmission lines, back-up and so on.
This article seems very balanced and informative and I would highly recommend it as a great starting point.
What is interesting is that while it attempts to include GHG's (CO2) and life cycle costs (very good so far) it is silent on the costs of NOx, SOx, Mercury, Particulate Matter or VOC (Volatile Organic Compounds) and so still tremendously UNDER-estimates the ACTUAL Total Cost of Energy which would in fact increase the cost of alternatives and make wind look far less expensive than it is already showing.
Also, nuclear decommissioning costs and taxpayer assumed insurance for catastrophic events have NOT been factored into the cost of nuclear. Recent costs for decommissioning in New Jersey (Constellation Energy) that are being paid by the local ratepayers is approximated at US$5Billion however I do not know at this moment how many MW are generated at this facility. It would seem that the calculation of Decommissioning Cost($)/MW should be relatively simple to undertake.
Also costs of water depletion are also not indicated for either fossil fuel generation types nor for nuclear. A recent IEA report showed that for each MW-Year of nuclear generation 1.4M gallons of water are consumed. This has a cost that is not charged to the nuclear company but is a significant cost to ecosystems and the economy. This should be charged at full industrial rates and then compared to what it would be at residential rates.
Furthermore, onshore wind has a cost of land and in many cases is on flat arable land and there are studies that show that wind farms on land can possibly create drier areas behind the windfarms that may lead to lower crop yields if there are crops or perhaps less vegetation for wildlife. No breakout has been made to differentiate offshore wind from onshore or near-shore or even aggregating onshore and near-shore together.
My guess is that if the above-noted adjustments were made that the comparative costs of fossil fuel and nuclear would be substantially higher than windpower and that offshore windpower would have the lowest costs of all energy forms.
I think the water is a bit of a nonissue; if you're sourcing it from the ocean it's effectively limitless and uses 2/3rds the surface of the Earth as the heatsink. Inland, build cooling towers, or larger artificial lakes, and run a closed loop.
You can't put seawater through a coal, nuclear etc plant, the salt stuffs everything up. You need fresh water.
You can of course use the plant's power to desalinate seawater to then use in the plant itself, but then you're getting less net energy from the thing.
Fresh water's a big issue for power plants. Here in Victoria we've had power prices rises largely because of that. Most of our power comes from hydroelectric and coal, both of which use a lot of water.
Actually it is possible to use seawater to cool nuclear powerplants (and I would assume other types as well).
I've visited Seabrook before - http://www.answers.com/topic/seabrook-station-nuclear-power-plant?cat=te...
So it can be done. It's not common, but Seabrook has a safe record now of 18 years of operation...
The Millstone nuclear powerplant in Waterford, CT, USA uses sea water as coolant. It's been in operation since 1970 without incident, although it was shut down for a year for a safety evaluation in the nineties. But the safety issues raised were not related to sea water as far as I know. My roommate in college worked there in the eighties and said that safety protocols were very lax (so he quit, returned to college, and became my roommate). It's currently in operation.
Here's an article in wikipedia: http://en.wikipedia.org/wiki/Millstone_Nuclear_Power_Plant
You can use seawater; it just doubles the investment because of the required materials. Also increases the required maintenance by a lot.
Cooling towers are not closed loop. A cooling tower system typically requires 5% of the water circulation as makeup. In a lot of areas even this amount of freshwater is unavailable; let alone a once-through system.
A true closed system is possible. Fin-fan units, ususally using a glycol mixture. It will cost a fortune though, which is why no one wants to do it.
Completely dry cooling systems that essentialy are large water to air heat exchangers are off-the-shelf systems and are for instance used for combined cycle gas turbine plants in deserts.
Also, nuclear decommissioning costs and taxpayer assumed insurance for catastrophic events have NOT been factored into the cost of nuclear. - Yosh Schmenge
Yosh, at present US utilities collect 0.1 to 0.2 cents/kWh to fund decommissioning. This funding is probably excessive since US reactors are now going through processes that extend their lives to 60 years, and research is underway aimed at extending their lives to 80 years. In the near term at least, it is probably cheaper to recondition old reactors to extend their lives than to replace them with new reactors. As for insurance, new reactor designs are extremely safe. The GE ESBWR will suffer a core meltdown once in every 29 million years of operation. If that once in every 29 million years accident takes place, safety back ups that are far superior to those in the Three Mile Island Reactor are in place. If you will recall, the Three Mile Island Reactor accident did not cause any deaths or injuries, and was not linked to any illnesses. In a rational world reactors should be insurable for catastrophic losses, and the premiums should not be expensive. If Insurance companies do not wish to offer risk coverage at a cost that is based on rational risk assessment, then it is in the interest of the government that insurance coverage be offered. Who ever offers insurance should have the right to charge a reasonable premium based on risk. The insurance premium would not be large in any case.
And computer hard drives have been consistently rated in millions of hours btw failures. However, I find that mine die after a couple of years of intermittent use.
When I hear such long term claims as that, I'm skeptical.
Reactors have numerous built in safety features. Reactor safety engineering is very sophisticated. Researchers have spent decades figuring out how reactors can break, and ways to prevent them breaking. This includes fool procedures, lit the automatic reactor shutdown in florida last week. At the same time researchers have figured it better ways to contain leaks in case an accident does happen. Before you describe yourself a skeptic, you should inform yourself about nuclear safety. If you proclaim yourself skeptical without having information, then you reveal yourself to be an ignoramus.
You'll be comforted in the fact that the hard drive manufacturers say the same thing.
They can say whatever they want, they dont go through the same safety regimen that nuclear engineering does. Thats bourne out in the safety record of nuclear power plants, which even including Chernobyl has the lowest death or injury count per GW/hr.
This 'well, my hard drive failed and they promised me reliability too' argument is specious nonsense.
If I've read the article correctly it seems to be saying gas (or hydro) peak plant can be phased out or at least reduced. Is that correct? Or put another way that wind and nuclear alone can supply average grid needs. If nuclear was designed to meet the average power draw (in case of windless conditions) but wind happened to be supplying 20% then current reactors would need to be quickly throttled back by that amount.
Current reactor designs are entirely unsuitable for this. The cost of nuclear fuel is nearly insignificant, so even if you could quickly throttle back a reactor its operating cost will barely budge. Additionally you don't want to put the fuel in a reactor through too many quick heating and cooling cycles.
This can easily be overcome if the "obsolate" energy is used for alternative purposes like electrolysis or pumped water storage. It may only reduce the financial appeal of the scheme slightly.
Also newer designs like PBR allow throttling back, it is just that it hardly ever makes sense in real world scenarios (which is the same situation as wind).
Nuclear for baseload and hydro for peak have worked nicely for decades in Sweden or France. I don't see a reason it can not be nuclear + hydro + wind + hydrogen +...
The cost of pumped storage is only slightly prohibitive? Please enlighten me.
I haven't looked at the cost of high temperature electrolysis plants, but even if you make lots of hydrogen gas, what are you going to do with it? It's still expensive to store and transport and even with high temperature electrolysis you're looking at some ~50% efficiency.
We're only 9 million people here in Sweden; we don't need a lot of swing capacity. Norway has only 4-5 million or so people. Between us it's some 180 TWh per year of hydro-electric, much of which goes to base-load(if you were to use it only for peak load I think you may need to install more turbines for the extra capacity). Without some kind of unprecedented expansion of the electric grid it will only ever be used to satisfy peak demands in Sweden, Norway, Denmark and to some extent Finland. Norway has some hydro electric resource not being exploited currently, France has about as much hydro-electric power as Sweden. That's the big ones, and it's still not close to enough; where are you going to find the rest of the hydro-electric resource?
For scale, the EU as a whole consumes 3000 TWh/year and that's set to rapidly increase as several member states and prospective member states were severely hurt by world war 2 and soviet occupation and are currently experiencing high levels of economic growth.
The most straightforward thing to do with hydrogen is to burn it in CCGT or use it in refineries. If burnt in CCGT the overall efficiency would probably be in the 40%, which makes it almost competitive with peaking power (peaking is usually priced ~2 times baseload). I have no idea what would be the cost of electrolysis plants, but it should be minimal as the equipment is relatively simple. Direct reforming into hydrogen in HTGR is also an option for the not too distant future.
I downplayed the need of pumped storage/hydrogen because in a good mix it would be minimal. How much pumped hydro does France have? Not that much compared to its massive nuclear fleet but it still has found its economical ways to use the excess power (export to Switzerland, load-leveling etc).
My point is that it has rarely been the excess power which causes problems because one can always disconnect units in seconds - for example problems with wind have rarely been related to power surges, it's the unexpected power drops and loss of reactive power that destabilize the grid.
Other options to balance wind/nuclear are IGCC (plus carbon capture?), solar thermal with storage, natural gas, waste, biomass etc.
My quarrel with Jerome has to do with two questions: What are our goals? And how do we best accomplish those goals? If our goal is to lower carbon emission in the generation of electricity, then wind has a place. Wind is a useful tool in lowering CO2 emissions. If, however, our goal is the elimination of CO2 emissions from the generation of electricity, wind becomes more of an obstacle than a help in meeting the goal.
I believe that the latter goal is far more desirable. If our over all goal is a 80% cut in global CO2 emissions by 2050, then the goal for the electrical generation sector should be zero emissions world wide. (I argue this because CO2 is easier to cut from electrical generation than from other economic sectors.) At the moment there are two candidate technologies for complete replacement of fossil fuel electrical generation. They are solar thermal and nuclear. Wind in not a candidate to replace fossil fuel generation, because it produces power too irregularly to serve as a basis for the power generating system, and for many electrical systems it becomes most unreliable during periods of peak demand.
A couple of years ago the staff or the Electrical Reliability Counsel of Texas argued that only two percent of the name plate generating capacity of Texas windmills should be counted as peak demand power. The reason was simple, during the periods when electrical demand will be at its absolute highest, that is during July and August daylight hours, when everyone in Texas is running their air conditioners full blast, the wind stops blowing all over Texas.
This means in practical terms that if the ERCOT system is going to meet 100% of the electrical demand at 2:00 PM on August 6th, it can rely on only virtually no electricity coming from wind powered sources. In a post carbon world only solar and nuclear generating sources can be relied on to produce electrical energy at two PM on August 6th in Texas.
The problem of summer wind reliability is by no means confined to Texas. It appears to be nation wide.
Therefore wind energy will start us down the road to to carbon free electricity, but it won't get us to the end of the road. If the post carbon electrical generating system is requires to generate virtually 100% of its peak electrical needs without wind, why have wind at all? The case for wind in the post carbon generating system is the weakest of all of the renewables.
I know that Jerome a Paris does not like to hear this, but investments in wind do not help us reach the post carbon world. Are they worth while as a temporary stopgap to lower CO2 emissions? I do not have an answer to that question, but if we have doubts about our ability to pay for a post carbon electrical system, why pay for the redundant and unreliable generating capacity of windmills, when more reliable post carbon generating systems are at hand?
I don't mind fact-based arguments like yours! You put on the table a very legitimate, and very real, issue that does require a serious answer.
This is worth a full post (and it's my intention to write it one of these days), but the simplified answer would be that wind can be (a large) part of the solution towards a post-carbon world, even with the issue you raise. In my view, the way to solve this is to keep the existing fleet of gas-fired plants in service, but put them in use only on those few days when they are needed. The overall gas consumption (and emissions) will be quite low - and certainly at tolerable levels altogether. I'd expect that you need to keep a chunk of gas-fired capacity in any case to provide peak capacity, and spinning reserves for flexibility and rapid response requirements of the network.
Other solutions will include beefing up the national high voltage network, to allow better connection between the existing different regions (a requirement even without wind considerations) and different wind regimes areas.
I wonder why solar has such high carbon emissions? Are they talking about rooftop solar (which is really competing with electricity distribution) or some kind of concentrating solar like all the projects being built these days?
I do know that the biggest solar plant complex now around is the Mojave desert one. That does have substantial carbon costs because it is a 24/7 power supply. It burns oil at night to keep the turbines running when the sun isn't shining and heating the heat transfer liquids that operate during the day.
I think mainly you are seeing northern european conditions there though there is also a lag as solar is improving its payback times pretty rapidly. The life span of a solar panel is pretty much set by exposure to cosmic rays which is 24/7 while the energy output is set by its exposure to the Sun which is low in Northern Europe. The energy input is set in large part by the energy needed to refine silicon and the amount of silicon used. Both are dropping pretty dramatically. Associated emissions estimates for wind and solar are based on the mix of energy generation on the grid, which is not usually the case for nucelar power which pretends that enrichment does not divert available power. This link on CSP was provided by a commentator on the Real Energy Blog.
http://www.ases.org/divisions/electric/newsletters/2006-04.html#roi
It suggests that CSP has a similar EROEI to wind and would thus have similar associated emissions since the inputs are similar: iron and concrete.
Chris
Simple: wind is the only thing which is politically, technically and economically feasible at this point of time. If I were to run these things I'd say that it's our version of "it's better than nothing", until we get the nukes going.
Where it becomes a drag and a dangerous proposition though (in tandem with solar PV and solar thermal without storage) is when they are sold as the main long-term tool for combating climate change and FF depletion. I seriously think that this myth is secretly pushed by the fossil fuel industry - ironically it is the gas and coal industry mostly interested in new wind! A 7-8% capacity credit means that 1MW of wind automatically needs 0.92-0.93MW of fossil fueled capacity running somewhere... for as long as the wind farm is running. And since wind farms are producing at most 30% of the time we will have ~0.3 wind vs ~0.9 fossil fuel continuous megawatts in the most perfect and ideal case - which gives a maximum of 25% of the electricity being wind vs 75% conventional in a closed grid. Therefore if a country like Germany goes ahead with its plans with closing its nukes it will have to replace them with coal - they simply have no other choice!
This is exactly why I have issues with the policies promoting renewable generation regardless of technical realities - I think they don't even grasp the amount of damage they are making for the long run. Wind is what it is - a good complementary energy source that can help us a lot in the medium term. Let's leave it there.
Wind still is a technology in search of a workable application. It seems attractive but how to make it valuable post fossil fuels?
One problems with a nuclear based system is how to accomodate the daily variations in demand. Hydro and gas are great solutions now, but what do we do as we phase out gas and cannot expand or maybe even lose some of our hydro.
We will have to find a way to daily build a despatchable reserve, perhaps through pumped storage. Maybe we run the nuclear above baseload with the excess building the despatchable reserve and feeding the intermittent wind into this process as well.
I think we will have to find a way to make wind and solar work in a power on demand grid where we do not have the current options for the variable demand, even is it costs quite a bit more per KWhr.
The reasons nuclear is run 24/7 are 90% economical and probably just 10% technical. There is zero incentative to invest in peaking nukes (and there are known designs that can do it like PBRs or some French PWRs) while 70% of per kwh costs are fixed before the plant has started up. For such an investments each minute the plant is not at max is a pure loss.
Personally I think that when nukes get streamlined in the next 2-3 decades the capital costs will drop so much that we will end up with peaking nukes too. Until then there are plenty of other peaking/balancing options and NG is not expected to run out that quickly.
Peaking nukes? That sure would be a problem for wind. I agree with you about dropping capital costs. There will also be a much more accommodating regulatory environment and more balanced requirement to pay for environmental damage that will strongly favor nuclear.
The NG might still be available but burning it for electricity seems such a waste. I am worried about being able to fill my propane tank.
If you turn your nukes on and off I will leave the vicinity. I wouldn't run even a gas cooled fast reactor any way except flat out, no throttle.
Dimensional changes are really, really, really, bad news. It's called metal fatigue and thermal shock.
I sort of doubt it. Peaking nukes for all purposes might as well be dumping excess load into resister banks to maintain constant output. While this is possible, it strikes me as a bit silly...
Levin, while I appreciate many of your other comments, I think you're repeating a myth here. For one 7-8 % capacity credit for wind is rare, usually it's in the range of 20-30% in the literature. The bigger myth you are putting forward is that wind needs conventional spinning reserve to cover the difference between nameplate and actual production. This is very far from the truth.
When the system operator runs unit commitment, he/she has to make sure there will be enough capacity during the next day and half to cover the demand. Operator has uncertainty how much demand will exactly be as well as uncertainty whether all plants will be able to deliver what they promise. Therefore there has to be reserves. Some of them are spinning or quickly adjustable hydro or load, since a power plant or transmission line can go offline in split second. However, wind power over large area does not change it's output fast (one wind farm might do that, but not wind power from multiple wind farms). The speed of change for wind power is such that it can be covered almost fully with non-spinning reserves. This means that the power plants that will take up production when wind falls off can be powered down while waiting for that to happen. Wind is reasonably predictable on typical unit commitment period of 36 hours and many markets offer nowadays possibilities for redoing unit commitments much closer to operating hour. Also many load following units are warm, but not running, and will fully startup only once called by system operator. They typically have 10-15 minutes to do it.
Another way to look at it: there's no reason to try to guarantee that wind plus something will provide wind power's nameplate capacity. It would be quite enough to guarantee that you will get at least the average production. This would mean that if wind farm is producing at 10% and the average production is 30%, one would need conventional capacity worth 20% of wind power plant's nameplate capacity to get the same capacity and energy contributions from wind. If you go all the way up to 100%, then you give much higher capacity contribution than energy contribution to the system. It would in practice be a peaking plant. Anyway, that's simplified and incorrect way to look at it, since it omits the rest of the system. This paragraph was just an attempt to highlight the mistake you are making.
Personally I think there are reasonable ways in the future to make the power system much more flexible and therefore to make it possible for wind or solar to be a big contributor to zero or low carbon energy system. Heating, cooling, transport, fuel production and manufacturing might offer much cheaper energy storage for electricity than conventional electricity storages, although some of that might also be economic. Still, it will be hard to beat the economics of a gas turbine in peaking operation.
I think there are reasonable ways in the future to make the power system much more flexible and therefore to make it possible for wind or solar to be a big contributor to zero or low carbon energy system. - JohnK
This flexibility may come with costs. A more flexible system may also be a more expensive system.
Certainly there will be costs. The question is whether those costs are lower than trying to achieve low carbon energy system with other means. One has to also take into account that wind power cost has not saturated yet and it's impossible to know how low solar PV costs might go. When those costs go lower, it will make more and more flexibility measures profitable compared to status quo. On the other hand fuel costs are on the rise for a good reason: scarcity. At some point this will include nuclear unless breeders or other fourth generation plants become feasible. Backstop price is probably in the cost of extracting uranium from seawater plus profit and risk margins. There is also the cost of getting rid of CO2 when using fossil fuel, which appears to be quite considerable plus that it will probably take about a decade before we know whether it'll work in large scale and at what price.
In the long term (wind saturation) overbuilding and grid linking (it's slightly inefficient, but so what?) to loadbalance solves most of your problem.
In the medium and long term, however, hydropower is an entirely throttleable resource (within whatever your environmental tolerances are). We don't necessarily need pumped hydro or serious grid linking to loadbalance a moderate load... we have normal hydro. With minimal work, you can create a system that accumulates water throughout the year and sends it downstream on non-windy weeks.
The amount of hydropower a nation uses gives you an immediate idea of how much windpower their grid can absorb without much modification.
You're right that wind is irregular in its output, but there is much work being done in firming variable renewable resources.
Next week I have a technical review of a presentation for a wind driven ammonia plant that will consume the output of a 230MW (faceplate) wind farm and produce 60ktons+ annually of ammonia. We primarily use this for fertilization here, but one of the people helping on the technical side of this is Dr. John Holbrook from the Ammonia Fuel Network and as a component of the plan we're looking at the operation of a 10MW ammonia powered "peaker" generating plant in conjunction with the production facility, which would allow us to take out a competing natural gas system. This will be formally presented to investors mid March and heard again at the first Iowa Wind Energy Conference near the end of the month.
That is ready today and the follow on work is even more exciting. Holbrook has a solid state ammonia synthesis process ready for pilot that takes about half the energy of the traditional H2 electrolysis => Haber Bosch and this works with gray water and produces no heat. The units involved are small and we're going to be doing some grant work associated with building a pilot involving a small scale wind turbine, perhaps around 600kw faceplate, an SSAS plant, and some sort of ammonia based electricity generation facility. There are fuel cell, piston engine, and turbine based options. I suspect there will have to be a SSAS/hydrolysis hybrid mode to this as the ammonia doesn't like to burn without a good, hot starter fuel like propane or hydrogen.
As soon as these things reach an acceptable state in the funding process we're going to turn our attention to railway right of way as wind farm/electric transmission corridor development. We've got a snippet of class 3 rail up here owned by a very inventive fellow and it happens to run right through one of two candidate sites for the wind driven ammonia plant ...
The world, it would appear to be screwed overall, but places with arable land and renewable resources have a chance. I hope I manage to actually facilitate the implementation of some of this stuff right here at home.
SacredCowTipper -
I think I understand your proposed concept of using a portion of the ammonia produced by you wind>electrolysis>ammonia scheme to serve as fuel for backup generation when wind power is insufficient.
However, I question whether burning ammonia is the most cost-effective way of providing that backup power. The reason I question that is because you are already making hydrogen via electolysis as feedstock for the ammonia plant. So, if you are already making the hydrogen, then why bother putting it through the ammonia plant and why not just store and burn a portion of the hydrogen directly?
I recognize that the long-term storage of hydrogen presents a number of problems, but we're not talking long-term storage here ...... perhaps only 24 or 48 hours worth.
Regardless of the merits of this particular scheme, I think that unless some form of short-term energy storage is eventually integrated into these large wind farms, it will always be unacceptable to have wind power constitute more than some relatively small fraction of a grid's total generating capacity.
An unpredictable, intermittent power supply is not merely inconvenient; it can be downright dangerous.
Addendum:
As an afterthought to my previous comments above, while using the hydrogen as a fuel might be better than going all the way to ammonia and then using the ammonia as a fuel (particularly if you can utilize the hydrogen in a fuel cell), neither option strikes me as a terrible efficient way of storing wind energy.
Taking the electricity you've gone through all the trouble to make and then converting it into chemical energy in the form of either ammonia or hydrogen, then releasing that chemical energy via combustion in a heat engine, then using that heat engine to turn a generator to (finally) make electricity strikes me as highly wasteful and conceptually inelegant.
Given the losses inherent in the electrolysis process coupled with the inherent losses of a heat engne plus generator, I'd venture that you would be able to store no more than 25% of the electricity generated by the wind turbine. Not good.
I maintain that if one is has already made electricity, then if at all possible, that electricity should be stored as electricity. Easier said than done, I know, but that should be the objective. Going back and forth like this is the thermodynamic equivalant of digging a hole, filling it back up, and then digging it again (or a gear train consisting of a gear with 40 teeth turning a second gear with 20 teeth, turning a third gear with 40 teeth, turning a fourth gear with 20 teeth ................. etc, etc.).
Charles, I wouldn't say winter was all that good in TX either. Just last week they came close to a blackout when their 1700MW of wind power suddenly dropped to 300MW. Spinning reserve and quick load shedding saved the day. That drop was about 4% of their 32,000MW total load. Without that spinning reserve they would have had units tripping all over the place. This is the problem with intermitten generation; you have to have quick reacting, such as gas turbines, units on line at part load to absord the wind farm/solar output fluctuations. These spinning reserve units are not high efficiency or low CO2. The spinning reserve must also be able to meet a large portion of the intermitten generation output. Jerome needs to add that to his bar chart for the wind generation.
Joe
Joe, I am quite aware of our recent wind event. The timing of such events during the winter is unpredictable, while it is predictable that wind speeds will drop all over Texas during many summer days when electricity demand peaks. We can can extend these predictions to other areas in the United States ands Canada, so the problem in not a local one in Texas. Obviously the advocates of a post-carbon role for wind generated electricity need to acknowledge the problem, study it, and come up with a reasonable plan, or the problem is going to make their business go away.
Texas event was a combination of things, not just wind. Wind did drop, but that's not a problem if it's predicted and this time it wasn't. It's part of normal power system operation to match production with demand. Wind just adds a new layer. Predictions are getting better, but it's hard to say how good they will be. In any case the event also included a 500 MW forecast error in demand and several committed conventional units, which failed to deliver. Couple of days earlier Florida had a blackout apparently due to substation fire, which forced two nuclear units to drop out. This was much worse situation than in Texas. That is not to say that wind will never cause blackouts. No power system is 100% certain to keep lights on at all times, because it would not be worth it. It would be overbuilding the system. When wind is added to the system, the system security actually increases unless load increases more than the capacity credit of the wind at the same time.
JohnK, Your second sentence says it all. Momentary wind output is not predictable. System load and capacity are always balanced, otherwise system frequency excursions will exceed acceptable limits and the generation units will start to disconnect and the system will go down. All non-intermitten generation units online will respond to load changes, if not at base load, on a momentary basis to maintain system frequency. Wind turbines cannot respond when more power is needed. While I agree that no power system is 100% reliable, we here in the US are lucky to have quite reliable electricity. Adding any intermmitten power units to the system cannot increase the system reliability without also adding sufficient spinning reserve to account for the momentary output fluctuations of that intemitten generation. Our partial blackout in FL was caused by human error, not by any equipment failure and yes, the system did NOT have sufficient spinning reserve to account for two nukes and two coal fired units, probably 3,000MW.
Large scale wind power does not do fast changes. Read my earlier comment. It took three hours for wind output to drop in Texas. That's plenty of time to start units, if enough capacity has been committed in the unit commitment.
I did not say that wind output is not predictable. You can do quite good predictions, but predictions have uncertainty. You have to take that uncertainty into account when doing unit commitment. It's same with load predictions, transmission failures and forced outages of conventional power plants. Some events are so unprobable that you rather take the risk that it happens than try to cover for it. In Texas couple of unprobable things happened and the result was that demand side management had to step in.
In some cases it might make sense to keep output from wind power plants below the amount they could produce from the wind they got. This is increasingly possible with modern turbine control systems. It might make sense in situations when the lost production is less valuable than getting the reserve capacity from some other source. At very large wind penetration levels there might be occasions that system security demands it. However, this will remain quite extraordinary measure, since usually it's cheaper to do it with conventional power plants.
Suggest you take a look at NREL/CP-500-30747; Short-Term Power Fluctuations of Large Wind Power Plants. Hundreds of turbines over large areas. 1 second, 1 minute, and 1 hour output data recorded over one year. The statistical analysis, unfortunately, uses nameplate rating rather than momentary output which makes the numbers look considerably better that if the output fluctuations were percent of output, not capacity. For the 12 month period, maximum 1 second fluctuation was +4.3% and -7.3%. For 1 minute, +11% and -14%. And for 1 hour, +63% and -50%. These percentages would be much larger if the variations were measured against the actual output at the time. While one sigma/two sigma variations were smaller, it is clear that backup generation will be of little use if not online at part load.
Charles says "Wind in not a candidate to replace fossil fuel generation.." then explains that "during July and August daylight hours....the wind stops blowing all over Texas". Well, maybe wind is not a candidate in Texas but that does not rule it out for everywhere else. In many other parts of the world it blows at the right time e.g. peak electricity may be needed for heating and lighting in the winter when the wind is stronger and the sun less. I am sure that there must be parts of the US where they have plenty of wind.
The surface winds of each hemisphere are divided into three wind belts at around 30 degrees for each belt. Texas straddles 30 degrees north, so wouldn't it generally be in between two belts and therefore not an obvious place for wind power:-)
There is no "one size fits all" each technology needs to be used where it is effective. It would be disingenuous to claim that PV could never work because say Ireland is always cloudy.
Toney, Actually Texas is considered to have good wind resources. The Stanford study which identified 17 locations with the most reliable winds included several locations in texas, and others in Oklahoma and New Mexico. I wont say that there are summer wind problems everywhere, but places where I have confirmed summer wind problems include, the provence of Ontario, California, the American South East where simmer capacity factors drop below 10%, New England, and the upper great planes. There are a few localities on the lower great planes where summer winds may be found.
"The problem of summer wind reliability is by no means confined to Texas. It appears to be nation wide."
I have lived in several areas of the US including the mountain west, northern plains and California. This blanket statement is not true for these areas. The wind often starts to blow stronger during summer sunny days due to the thermal effect. And in fact the plains states often have wind blowing for long periods; witness Fargo, ND where during weather extremes the wind usually blows steadily at 10 to 20mph for two or three days.
mbnewtrain, "The Wind Energy Resource Atlas of the United States states,
In summer, wind speeds aloft diminish, and wind power is at its lowest over most of the United States. Although only class 1 or 2 wind power occurs over much of the contiguous United States."
http://rredc.nrel.gov/wind/pubs/atlas/chp2.html
It goes on to identify several areas where wind resources are considered Class 3 or better. These would include California, New England and the upper great planes. But other findings suggests that parts of this report may be in error. For example, during the California "heat storm" of 2006 the thermal effect failed, and California winds died away to almost nothing. According to David Dixon, between "August 17 to 23, wind produced at 89.4 to 113.0 MW, averaging only 99.1 MW at the time of peak demand or just 4% of rated capacity.
http://www.energypulse.net/centers/article/article_display.cfm?a_id=1332
Stephen R. Connors of MIT analized data from the National Oceanic and Atmospheric Administration (NOAA), U.S. Environmental Protection Agency, and other data sources tracking wind speeds, electricity demand, and power systems operations over time. Conner found that offshore wind farms in the Northeast would generate far more electricity in winter than in summer, when electricity demand is highest. However, Connor noted about New England off shore wind power that "it’s not there when you need it most—midday during the summer when demand and prices are high,” Never the less, Connor believes that wind power plays a useful role in reducing CO2 emissions.
http://esd.mit.edu/esd_reports/winter2006/offshore_wind_power.html
I do have a quote from Mike Cowan of Western Area Power Administration’s Upper Great Plains Region, but the link has failed.
The Atlas also claims good summer wind resources for the Texas coast, but Texas coastal wind generators Pose a threat to migrating birds:
http://www.prnewsnow.com/Public_Release/Government/174882.html
In addition the problem of Texas summer wind unreliability extends to coastal generating facilities as well. When the ERCOT staff considered the reliability of Texas winds, the considered the reliability of coastal winds as well. They came to the conclusion that only 2% of nameplate Texas wind generating capacity was sufficiently reliable to be included as peak electricity generating capacity.
Advocates of wind energy need to be very careful about making reliability claims. One day ill considered claims can come back to haunt the business.
On the Pacific coast of California, at the hottest times, the land-sea temperature differential is the largest so the wind blows the strongest.
Except during the peaks when the hot onshore air pushes the marine layer offshore. Then the air is dead. That can go on for a week or more and will only get worse as the climate heats up, at least here in the Bay Area.
Isn't it possible that wind could be used as a baseload if a north-american-wide network of wind turbines from Texas in the south to Alberta in the north were created? The point being that the wind is always blowing somewhere on the continent. I understand that you'd have incredible start up costs to create something like this, installing tens of thousands of turbines at different latitudes. Has anyone done the analysis for something like this?
Yeah, lots of studies.
Here are a couple of links:
http://www.economist.com/science/displaystory.cfm?story_id=9539765
Power transmission | Where the wind blows | Economist.com
http://www.sciencedaily.com/releases/2007/10/071019184844.htm
And here is a less optimistic view:
http://www.windaction.org/documents/461
www.windaction.org | E.ON Netz Wind Report 2005
Since wind power itself requires large subsidies before you start budgeting for massive power grids, most of the super grid advocates seem to have the natural optimism of someone intending to spend someone else's money to me, but YMMV
the wind is always blowing somewhere - Arclite
This is the fundamental myth of the wind crowd. In fact it is possible for winds to be insufficient everywhere to meet high summer power demands.
You could compliment with solar.
You could indeed - so long as the demand did not occur at night, or maybe during the winter, when even at the latitude of the Mohave you only get around 25% of the sunshine you get in midsummer - shorter days and so on.
I have great hopes for solar, but at the moment they remain just that, hopes.
Solar energy is far more immature than wind power, proposals and a couple of builds are underway at the moment to see how we can do, especially with solar thermal, but it is challenging at this time just to provide peak power in hot areas where the main demand is for cooling, and indeed it seems likely that for this use at least solar should do the job, and so you are correct, but at the moment only in some limited areas.
Proposals to run substantial parts of our grid on solar are not really serious engineering at the moment, but are based on hypothetical massive advances in cost and technology.
We know pretty well, OTOH, how to run up to maybe 30% of the grid on wind power, but it would be somewhere between very expensive and fantastic in cost, you would need a fair amount of back-up although by no means one to one, (Denmark relies on scads of cheap hydroelectric from Scandanavia ) and you would build in fossil fuel burn to make up for intermittencies.
Once item that must be considered it load response time - even with a base load of nuclear, and a large amount of wind generation in place, there will still be need for gas fired power plants to cover short term generation gaps as wind drops for a couple of minutes.
It is the relatively slow response from coal power plants that will be easy to displace, but they currently hold the low marginal cost/base-load capacity within our generation network. Nuclear has an even longer response time than coal, and the output from wind is variable, and at the mercy of nature. - so i would guess that there will be a place for hydrocarbon based generation for quite a while yet, as there is no easy substitute for quick response, high capacity generation (other than hydro, and there isn't sufficient capacity around in many places)
It's absolutely true. I do not know any power grid without a minimal basis of thermal or hydroelectrical plants. Consider also that building a nuclear plant or a windmill requires a minimal amount of fossil fuels - already to make the steel and the concrete. Actually all alternatives are fossil energy multipliers, but can not substitute them entirely. And when fossiles will be VERY expensive, their building costs of everythinh will ALSO be very expensive. Until I see a country whose industrial development is based on wind or nuclear energy, I won't believe they can become a reliable basis to energize the world by themselves...
Paul Gipe is doing a good bit of work on pushing ART - advanced renewable tariffs. Details are in the link below and he craves endorsements, both from individuals and industry/NGO contacts ...
http://wind-works.org/articles/feed_laws.html
You could argue that if feed-in tariffs increase the retail price of electricity they are a tax on the poor. The pensioner brewing a mug of hot chocolate doesn't have the chance of breaking even by feeding back electricity to the grid. On the other hand you could say power was going to get expensive anyway so learn to conserve now.
An analysis should compare feed-in tariffs with aggressive carbon caps. The extra rise in fossil burn costs will create a price gap for clean tech to expand into. It also ensures that clean actually displaces dirty, not just adds to it. To wit Germany's generous feed-in tariff for solar PV (is it 45 eurocents per kwh?) hasn't stopped talk of new coal stations.
Jerome, the figures I have seen for cost do not even come close to economic for wind turbines.
There is a proposal to build 33GW of close off-shore wind power in the UK. This would give according to Government figures around 10-11GW of actual average power output, and will cost £66bn.
At over £6bn per GW, and that does not include the extra costs of running back-up power or beefing up the grid, then it is probably somewhere between 2-4 times as expensive as generating the same power from Gen111+ nuclear reactors.
In Britain, wind power has excellent load-following characteristics, being around 2.5 times as powerful in the winter as in the summer.
Even then, that is not remotely close to being economic.
Britain has some of the best wind resources in Europe, and although construction costs on land are around half of off-shore, that is not hopeful for truly economic wind-power generation in Europe.
Charles has already indicated some of the problems in parts of the US, as many places there are hot and need a lot of power for air conditioning in the dog days of summer.
Prospects in India and China may be a lot brighter, as they need all the power they can get and construction costs are much lower there, but to my mind the main thing wind-power will do in most places in the West would be to lock in fossil fuel burn as a back-up.
There may be exceptions in areas such as Chicago.
Jerome said: